What's new
What's new

Matching RPM on multiple DC gearmotors

calderp

Aluminum
Joined
Jul 22, 2011
Location
New Orleans
I'm designing something that will have 2-3 components running simultaneously and they will need to be variable speed, reversible, and have fairly precisely matched speeds (within say 2%). I have done plenty with DC gearmotors and speed controllers but I've never had an application where I needed to match speeds like this. If I have to I can run them off a single motor and mechanically connect the components but it would be ideal to run three separate motors if I can get them to match RPM. Loads are not huge but I think more than I would expect from stepper motors and I would love to avoid programming anyway. Probably in the neighborhood of 100-200 in-lbs. Is there an off-the-shelf solution to this? Something similar to these constant-speed motors from Mcmaster, but in a gearmotor? McMaster-Carr
Obviously the different loads on the different components would result in a difference in speed were I to just drive the motors off of the same speed control. Is there an easy way to create this feedback loop where voltage is varied to achieve a desired RPM or is this simply not an application for PMDC gearmotors?
 
Do you need to control just speed, or both speed and phase? I.e., if one motor slips a little, is it sufficient to get it up to the target speed or does it also need to make back the angular slip?
 
You need new 3 high end dc drives controlled by 0 to 10 volt control with tack feed back or 4-20 ma control with tack feed back, and a master control to gang all 3 motors (or more)...Not cheap...vfds and ac servo drives would be better...Phil
 
Do you need to control just speed, or both speed and phase? I.e., if one motor slips a little, is it sufficient to get it up to the target speed or does it also need to make back the angular slip?

Some slip is ok, I think. You're a little over my head there but I think I follow. I don't need the linear travel of the systems to be matched exactly as long as the RPMs (and therefore linear travel speeds) sync once up to speed if there's a little slip when starting that should be ok. At the end of each cycle we could bump one or the other motor to line them up again.

All travel will be linear, one system is is essence a treadmill/conveyor and the other is a pair of cable or chain driven tracks similar to an overhead conveyor. They will be run, then stopped, reversed to starting position, run again. They will also likely need soft start and stop.
 
You can do the catch up with home pos limits, Look for drives that work with 4-20 ma control and use one master control, that's how I control a beer bottle line...Phil
 
Is it hard-locked? Or can you get away with "dancer" setups to take up material and trigger speed changes?
 
Use AC tachometers. I can send you a circuit for synchronizing two AC generators. The nice thing about this circuit is that it locks them with zero degrees phase angle where a lot of phase locked loops lock 90 degrees apart. Pick one motor for master and lock the others to it.

You could also use one motor and two AC synchros. Either of these approaches will let you soft start one motor and the others will track with it.

Bill
 
These guys have some awesome drives.
Minarik Drives
They'll do just what you want, if you have a fat wallet. I've used a half-dozen of their 2hp regenerative drives, with and without PWM. Either way they have great speed regulation if you use tachometer feedback like Phil and others have suggested. Even using armature feedback you might get good enough results if the motors are all the same and the loads are similar too.
 
I'm designing something that will have 2-3 components running simultaneously and they will need to be variable speed, reversible, and have fairly precisely matched speeds (within say 2%). I have done plenty with DC gearmotors and speed controllers but I've never had an application where I needed to match speeds like this. If I have to I can run them off a single motor and mechanically connect the components but it would be ideal to run three separate motors if I can get them to match RPM. Loads are not huge but I think more than I would expect from stepper motors and I would love to avoid programming anyway. Probably in the neighborhood of 100-200 in-lbs. Is there an off-the-shelf solution to this? Something similar to these constant-speed motors from Mcmaster, but in a gearmotor? McMaster-Carr
Obviously the different loads on the different components would result in a difference in speed were I to just drive the motors off of the same speed control. Is there an easy way to create this feedback loop where voltage is varied to achieve a desired RPM or is this simply not an application for PMDC gearmotors?

Dirt-common need. Regulation can be better than 1/10 of 1% if you need it.

Use the onboard or plug-in modules the major industrial-grade DC drive makers provide for networked control & sync, "DIY" nor hobby-grade goods neither required nor recommended.

I'd suggest the fibre-optic links if it is an electrically noisy environment.

Stock off-the self parts, "how to" is in the maker's manuals, and all have been in use long enough - 40 years and more - there's a lot of the gear in the used market for decades already if you need to do a proof-of-concept on the cheap.

If you don't NEED DC's ability to DELIVER, REGARDLESS, not slip, to high precision, and AC is good enough?

AC can be a great deal less expensive. Just use cheap and cheerful higher-rated motors so you have enough reserve they don't even get into high slip load ranges. Trade more spend on space for less spend in money, IOW.

Good DC equipment, and compact, can easily cost ten times what "good enough" AC costs.

Industrial grade VFD's are set up with options for the same links for networked control and sync over Copper or fiber as DC drives are.

Ethernet and WiFi networking for AC or DC also exist.

Some of the applications have wider span than others.

:D
 
Dirt-common need. Regulation can be better than 1/10 of 1% if you need it.

Use the onboard or plug-in modules the major industrial-grade DC drive makers provide for networked control & sync, "DIY" nor hobby-grade goods neither required nor recommended.

I'd suggest the fibre-optic links if it is an electrically noisy environment.

Stock off-the self parts, "how to" is in the maker's manuals, and all have been in use long enough - 40 years and more - there's a lot of the gear in the used market for decades already if you need to do a proof-of-concept on the cheap.

If you don't NEED DC's ability to DELIVER, REGARDLESS, not slip, to high precision, and AC is good enough?

AC can be a great deal less expensive. Just use cheap and cheerful higher-rated motors so you have enough reserve they don't even get into high slip load ranges. Trade more spend on space for less spend in money, IOW.

Good DC equipment, and compact, can easily cost ten times what "good enough" AC costs.

Industrial grade VFD's are set up with options for the same links for networked control and sync over Copper or fiber as DC drives are.

Ethernet and WiFi networking for AC or DC also exist.

Some of the applications have wider span than others.

:D

Thanks. Do know a manufacturer off the top off your head that makes these DC drives? I think that I can work up to $5k into the budget for motors and controllers, does that seem way off-base for synchronized DC PWM? My concern with AC is the speeds I'll be working with, looking at linear travel in a range of a few in/s at the low end and maybe 2fps max. Of course I could gear down but the simplicity and compact size of DC gearmotors would be ideal.
 
These guys have some awesome drives.
Minarik Drives
They'll do just what you want, if you have a fat wallet. I've used a half-dozen of their 2hp regenerative drives, with and without PWM. Either way they have great speed regulation if you use tachometer feedback like Phil and others have suggested. Even using armature feedback you might get good enough results if the motors are all the same and the loads are similar too.

Thanks I will call them. Do you think that $5k for 2-3 small gearmotors with controllers is in the ballpark? I'm fairly confident armature feedback would work for me.
 
Use AC tachometers. I can send you a circuit for synchronizing two AC generators. The nice thing about this circuit is that it locks them with zero degrees phase angle where a lot of phase locked loops lock 90 degrees apart. Pick one motor for master and lock the others to it.

You could also use one motor and two AC synchros. Either of these approaches will let you soft start one motor and the others will track with it.

Bill

Can I get slow and variable speeds with reversing though?
 
You can do the catch up with home pos limits, Look for drives that work with 4-20 ma control and use one master control, that's how I control a beer bottle line...Phil

Thanks Phil, it sounds like you're very familiar with that I'm trying to do. You're recommending AC servos and VFDs, does that still give me the ability to have very slow travel speeds and reversing? Could I run a 3-phase AC gearmotor with VFD controller to get a similar effect to DC PWM? Sorry for the basic questions, AC motion control is very new to me. Do you think that ~$5k could be enough for 2-3 DC motors and controllers and if so is there a manufacturer you recommend? I'm having trouble googling stuff on 4-20ms control and master controllers
 
Ok, I've been doing a lot of reading and I'm starting to getter a better handle on this. Appreciate this info pointing me in the right direction. It does seem like I should be looking into 3-phase AC motors with VFD controls so that's where I'm focused at the moment.
 

Attachments

  • Screenshot (38).jpg
    Screenshot (38).jpg
    21 KB · Views: 18
Last edited:
Are you trying to doo some kind of "Web" application ?
Where you have a long strip of material going thru a process ?
 
Are you trying to doo some kind of "Web" application ?
Where you have a long strip of material going thru a process ?

No, pretty weird application it's for a visual effect actually, not an industrial application. Essentially there will be a pair of 250' long chain-driven overhead tracks pulling about 100 linear feet of some vertical elements on each side and in center there will be a roughly 6' x 10' conveyor fixed to the floor and rotating in place. I need to the vertical elements to travel past the conveyor at the same speed and I'm trying to avoid having to run mechanical power transmission from the conveyor shaft to the overhead track. It will run a complete cycle in one direction, then reverse to starting position and repeat, at varying speeds. The cycle won't be automated other than limit stops for safety, we just need a speed control (analog or digital), a start/stop, and a fwd/rev. No need for linear controls or feedback or programming cycles or anything like that the only concern is matching the speed of the componenets. So it's very very simple programing the thing that's tripping me up is finding an affordable solution to matching RPM on multiple motors which are moving a different load. Also the vertical elements will be changed at points so I don't think we can use something really basic like IR compensation to match the motor speeds because the load may change and I don't want to have to re-calibrate. The ideal solution is a simple speed control which gives true RPM adjustments not just voltage (DC) or frequency (AC) adjustments.
 
Can I get slow and variable speeds with reversing though?

You would want to use motors with encoders. You would run one motor with normal DC power and the other two would run in lockstep. I have not run this circuit at very slow speeds. It looks at pulses from the two encoders. Motor A sets the circuit. If two reset pulses come before the second set pulse, it raises the voltage on the slave motor. If it gets two set pulses come before the second reset pulse, it lowers the voltage. I would expect that if the master motor is running and the slave motor stopped, then turned on, the slave would accelerate to match speeds, but would have made a different number of revolutions. If you turned the power on both motors at the same time, that difference should be minor. The encoders can be simple, not two phase with home markers like the ones on CNC machines. You could reverse with a relay.

If you can afford it, I would suggest using a motion control card and run it like a CNC machine where you can count pulses and keep the motors in synch. If something delays one of the motors, it will catch up. I do that with an Aerotech motion control card in a PC. That is probably more than you want to get into, but you probably could do it with a couple of Arduinos and simple transistor drivers.

Baldor makes gear head motors with encoders. I checked, but their web site is down, no doubt temporary.

Bill
 
You would want to use motors with encoders. You would run one motor with normal DC power and the other two would run in lockstep. I have not run this circuit at very slow speeds. It looks at pulses from the two encoders. Motor A sets the circuit. If two reset pulses come before the second set pulse, it raises the voltage on the slave motor. If it gets two set pulses come before the second reset pulse, it lowers the voltage. I would expect that if the master motor is running and the slave motor stopped, then turned on, the slave would accelerate to match speeds, but would have made a different number of revolutions. If you turned the power on both motors at the same time, that difference should be minor. The encoders can be simple, not two phase with home markers like the ones on CNC machines. You could reverse with a relay.

If you can afford it, I would suggest using a motion control card and run it like a CNC machine where you can count pulses and keep the motors in synch. If something delays one of the motors, it will catch up. I do that with an Aerotech motion control card in a PC. That is probably more than you want to get into, but you probably could do it with a couple of Arduinos and simple transistor drivers.

Baldor makes gear head motors with encoders. I checked, but their web site is down, no doubt temporary.

Bill

Thank you very much Bill this is very helpful! I will look into this tomorrow. Unfortunately I think I need an off-the-shelf option but maybe I can get that from Baldor. I just don't have the time on this project to figure out the programming side of things. In terms of the matching there isn't too much concern there, I don't need a perfectly synced linear travel only a synced speed, ie. there will be a built in startup period where the motors could 'catch up' to one another.
 








 
Back
Top